Re: Rack populating??-Rationalize it!!
When the radiating or receiving source (wire or antenna) is a significant
fraction of the separation between wire and antenna or antenna and antenna,
then not only are the (1/r to the n) calculations imprecise, they are WRONG.
All those 1/r to the n equations are based on a simplification that ignores
the length of the radiating element relative to the separation between
radiator and receptor. Therefore, the equations break down when this
situation is not obtained in reality.
--
From: CARTER car...@amcomm.com
To: 'Ken Javor' ken.ja...@emccompliance.com, CARTER car...@amcomm.com,
CARTER car...@amcomm.com, 'jim_bac...@mail.monarch.com'
jim_bac...@mail.monarch.com, franz gisin emc_...@yahoo.com,
emc-p...@ieee.org
Subject: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 10:38 AM
I agree that the extrapolations are not, shall we say, extremely precise.
Nor are most of the measurements we do - regardless of our confidence in our
equipment and our expertise. All the more reason to try to make test
environments repeatable, and EUT's configured for maximized (worst case?)
emissions.
-Original Message-
From: Ken Javor [mailto:ken.ja...@emccompliance.com]
Sent: Friday, July 28, 2000 3:40 PM
To: CARTER; carter; 'jim_bac...@mail.monarch.com'; franz gisin;
emc-p...@ieee.org
Subject: Re: Rack populating??-Rationalize it!!
--
From: CARTER car...@amcomm.com
To: 'Ken Javor' ken.ja...@emccompliance.com, CARTER
car...@amcomm.com,
'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz Gisin
emc_...@yahoo.com, emc-p...@ieee.org
Subject: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 10:14 AM
I have long forgotten the equation for computing free space transmission
loss, but it seems to me that when radiated power, frequency, and distance
are known, one can certainly predict field intensity at any other
distance.
If that is not true, we have made a large number of terrestrial microwave
transmission paths work purely by accident.
MICROWAVE LINKS WORK IN THE FAR FIELD.
Also, CISPR 22 11.2.1:
NOTE - If the field-strength measurement at 10 m cannot be made because
of
high ambient noise levels or for other reasons, measurements of Class B
EUT's may be made at a closer distance, for example 3 m. An inverse
proportionality factor of 20 dB per decade should be used to normalize the
measured data to the specified distance for determining compliance. Card
should be taken in the measurement of large EUT's at 3 m at frequencies
near
30 MHz due to near field effects.
10 METER TO 3 METER SCALING WORKS ONLY WHEN THE EUT IS SMALL WITH RESPECT TO
3 METERS, AND WHEN THE ANTENNA DIMESNIOS ARE SMALL WITH RESPECT TO 3 M. IF
YOU TRIED TO DO A 3 METER SITE ATTENUATION WITH TWO 30 MHz TUNED DIPOLES,
YOU WOULD FIND YOU NEEDED A CORRECTION FACTOR TO ACCOUNT FOR THE NEAR FIELD
EFECTS.
Someone at IEC thinks its possible. And many labs do, in fact, test at 3
meters.
THERE HAVE BEEN A LOT OF PAPERS WRITTEN SAYING THE EXTRAPOLATION IS FAR FROM
ACCURATE.
I was not suggesting that it would be wise or meaningful to do these
things.
My point is simply that it is wise to maximize the emissions from an EUT
to
insure that we are not missing emissions which may be above the limits,
and
that the end user of the equipment is not likely to inadvertently create
such a situation.
UNDERSTAND.
_\\|//_
(' O-O ')
ooO-(_)-Ooo
Mark Carter
AM Communications, Inc.
car...@amcomm.com mailto:car...@amcomm.com
Voice: 215-538-8710
Fax: 215-538-8779
-Original Message-
From: Ken Javor [mailto:ken.ja...@emccompliance.com]
Sent: Friday, July 28, 2000 2:51 PM
To: CARTER; 'jim_bac...@mail.monarch.com'; Franz Gisin;
emc-p...@ieee.org
Subject: Re: Rack populating??-Rationalize it!!
There is a very big, very important error in Mr. Carter's point number 2.
You absolutely CANNOT extrapolate from 10 m to any other distance, unless
that other distance also happens to be both in the far field of the EUT
AND
the far field of the measurement antenna. That is the reason for a 10 m
measurement in the first place. Everyone would be doing three meter
measurements if it weren't for issues related to three meters not being
far
enough away.
Extrapolating a 10 m measurement to one airline seat away is totally
impossible. And completely unnecessary: the victim protected by the 10 m
measurement is not some arbitrary gizmoid, it is a radio receiver. You
aren't supposed to be operating a radio receiver on the airplane. More to
the point, the problem with laptops and other personal electronics on a
commercial transport is not EMI to other personal electronics, but
interference with aircraft antenna-connected receivers, whose antennas are
mounted external to the aircraft.
--
From: CARTER car...@amcomm.com
To: 'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz
Gisin
emc_...@yahoo.com, emc-p
RE: Rack populating??-Rationalize it!!
Two EMC issues were discussed how to populate rack for radiated emission test and correlation of radiated emission tests at different test distances. For some EUTs there are too many possible mix and match configurations, with different or same modules, that testing all of them in order to identify the worst would be practically impossible. There is also no guarantee that full rack will be worse than less populated rack. Switching power supplies under different loads might have different EMC profiles. Cavity created by empty slots could resonate and increase certain emission levels. There is no good way to identify worst case configuration. But reasonable effort should be there and procedure to maintain test consistency and repeatability. CISPR SC G developed 2 dB procedure to determine representative ITE configuration for EUTs with multiple ports and/or modules of the same type. Procedure is in Third Edition CISPR 22:1997 Section 8.1 EUT Configuration. CISPR 11:1997 has procedure from Second Edition CISPR 22:1993 Section 9.1 which requires one module/cable of each type. Reference test distance is test distance at which limit is specified. Other test distances could be used, but test at the reference distance would take precedence. Manufacturer who decide to test product at other than reference test distance will assume the risk of poor correlation of test results, resulting in costly over or under EMC design. Potential customer or enforcement agency could re-test EUT and decide on further actions based on test data at the reference distance. My 2 cents... Mirko Matejic --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: Rack populating??-Rationalize it!!
--
From: CARTER car...@amcomm.com
To: 'Ken Javor' ken.ja...@emccompliance.com, CARTER car...@amcomm.com,
'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz Gisin
emc_...@yahoo.com, emc-p...@ieee.org
Subject: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 10:14 AM
I have long forgotten the equation for computing free space transmission
loss, but it seems to me that when radiated power, frequency, and distance
are known, one can certainly predict field intensity at any other distance.
If that is not true, we have made a large number of terrestrial microwave
transmission paths work purely by accident.
MICROWAVE LINKS WORK IN THE FAR FIELD.
Also, CISPR 22 11.2.1:
NOTE - If the field-strength measurement at 10 m cannot be made because of
high ambient noise levels or for other reasons, measurements of Class B
EUT's may be made at a closer distance, for example 3 m. An inverse
proportionality factor of 20 dB per decade should be used to normalize the
measured data to the specified distance for determining compliance. Card
should be taken in the measurement of large EUT's at 3 m at frequencies near
30 MHz due to near field effects.
10 METER TO 3 METER SCALING WORKS ONLY WHEN THE EUT IS SMALL WITH RESPECT TO
3 METERS, AND WHEN THE ANTENNA DIMESNIOS ARE SMALL WITH RESPECT TO 3 M. IF
YOU TRIED TO DO A 3 METER SITE ATTENUATION WITH TWO 30 MHz TUNED DIPOLES,
YOU WOULD FIND YOU NEEDED A CORRECTION FACTOR TO ACCOUNT FOR THE NEAR FIELD
EFECTS.
Someone at IEC thinks its possible. And many labs do, in fact, test at 3
meters.
THERE HAVE BEEN A LOT OF PAPERS WRITTEN SAYING THE EXTRAPOLATION IS FAR FROM
ACCURATE.
I was not suggesting that it would be wise or meaningful to do these things.
My point is simply that it is wise to maximize the emissions from an EUT to
insure that we are not missing emissions which may be above the limits, and
that the end user of the equipment is not likely to inadvertently create
such a situation.
UNDERSTAND.
_\\|//_
(' O-O ')
ooO-(_)-Ooo
Mark Carter
AM Communications, Inc.
car...@amcomm.com mailto:car...@amcomm.com
Voice: 215-538-8710
Fax: 215-538-8779
-Original Message-
From: Ken Javor [mailto:ken.ja...@emccompliance.com]
Sent: Friday, July 28, 2000 2:51 PM
To: CARTER; 'jim_bac...@mail.monarch.com'; Franz Gisin;
emc-p...@ieee.org
Subject: Re: Rack populating??-Rationalize it!!
There is a very big, very important error in Mr. Carter's point number 2.
You absolutely CANNOT extrapolate from 10 m to any other distance, unless
that other distance also happens to be both in the far field of the EUT AND
the far field of the measurement antenna. That is the reason for a 10 m
measurement in the first place. Everyone would be doing three meter
measurements if it weren't for issues related to three meters not being far
enough away.
Extrapolating a 10 m measurement to one airline seat away is totally
impossible. And completely unnecessary: the victim protected by the 10 m
measurement is not some arbitrary gizmoid, it is a radio receiver. You
aren't supposed to be operating a radio receiver on the airplane. More to
the point, the problem with laptops and other personal electronics on a
commercial transport is not EMI to other personal electronics, but
interference with aircraft antenna-connected receivers, whose antennas are
mounted external to the aircraft.
--
From: CARTER car...@amcomm.com
To: 'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz
Gisin
emc_...@yahoo.com, emc-p...@ieee.org
Subject: RE: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 7:35 AM
Jim,
I differ in opinion on at least two counts:
1. CISPR 22 9.1 states An attempt shall be made to maximize the
disturbance
consistent with typical applications . . . etc. This can and should be
construed as creating a worst-case scenario. Worst case is always assumed
to
be within the range of operating conditions reasonably expected. If a
laptop
will have higher emissions when it is in flames, no one (at least no one I
know) is suggesting that it be tested that way. Setting a laptop ablaze is
not within the range of reason.
2. The intent of making measurements at 10 meters, or 3 meters, or with an
absorbing clamp, or any of the other requirements is not to insure that no
interference will occur at 10 meters, 3 meters, or in a clamp, but that
test
results are repeatable. If the limit is 40 dBuV/m at 10 meters, and I test
at 20 meters, will the EUT pass? Of course not. If I have test results
from
a 10 meter set-up, I can compute the effect of an emission at any other
distance from the EUT to the next airline seat, or whatever. If I wanted
to
know that. The point is that we're all on the same page that way.
If you do the math, you can compute the interfering field strength of an
emission from a device in the next airline seat, and if you know the
immunity
Re: Rack populating??-Rationalize it!!
There is a very big, very important error in Mr. Carter's point number 2. You absolutely CANNOT extrapolate from 10 m to any other distance, unless that other distance also happens to be both in the far field of the EUT AND the far field of the measurement antenna. That is the reason for a 10 m measurement in the first place. Everyone would be doing three meter measurements if it weren't for issues related to three meters not being far enough away. Extrapolating a 10 m measurement to one airline seat away is totally impossible. And completely unnecessary: the victim protected by the 10 m measurement is not some arbitrary gizmoid, it is a radio receiver. You aren't supposed to be operating a radio receiver on the airplane. More to the point, the problem with laptops and other personal electronics on a commercial transport is not EMI to other personal electronics, but interference with aircraft antenna-connected receivers, whose antennas are mounted external to the aircraft. -- From: CARTER car...@amcomm.com To: 'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz Gisin emc_...@yahoo.com, emc-p...@ieee.org Subject: RE: RE: Rack populating??-Rationalize it!! Date: Fri, Jul 28, 2000, 7:35 AM Jim, I differ in opinion on at least two counts: 1. CISPR 22 9.1 states An attempt shall be made to maximize the disturbance consistent with typical applications . . . etc. This can and should be construed as creating a worst-case scenario. Worst case is always assumed to be within the range of operating conditions reasonably expected. If a laptop will have higher emissions when it is in flames, no one (at least no one I know) is suggesting that it be tested that way. Setting a laptop ablaze is not within the range of reason. 2. The intent of making measurements at 10 meters, or 3 meters, or with an absorbing clamp, or any of the other requirements is not to insure that no interference will occur at 10 meters, 3 meters, or in a clamp, but that test results are repeatable. If the limit is 40 dBuV/m at 10 meters, and I test at 20 meters, will the EUT pass? Of course not. If I have test results from a 10 meter set-up, I can compute the effect of an emission at any other distance from the EUT to the next airline seat, or whatever. If I wanted to know that. The point is that we're all on the same page that way. If you do the math, you can compute the interfering field strength of an emission from a device in the next airline seat, and if you know the immunity of the gizmoid in that seat, you can pretty much be certain that will not be any disruptive interference. My 2 cents -Original Message- From: jim_bac...@mail.monarch.com [mailto:jim_bac...@mail.monarch.com] Sent: Friday, July 28, 2000 10:36 AM To: Franz Gisin; emc-p...@ieee.org Subject: Re:RE: Rack populating??-Rationalize it!! forwarding for Franz Reply Separator Subject:RE: Rack populating??-Rationalize it!! Author: Franz Gisin emc_...@yahoo.com Date: 7/27/00 6:48 PM Worst case is not relevant when it comes to defining EMC test configurations for ITE equipment. I do not know of any ITE EMC test standard that specifically states worst case must be used. Eveywhere I look I see words like typical or representative or minimum. I am willing to bet that 99.9% of you, when you bought your last car, did not insist the car be at least 6 dB under the smog limits under worst case conditions before you bought it (e.g. ask the car manufacturer to load the car down with bricks, take it to a very steep hill, facing the front of the car up the hill instead of down, and then pushed the accelerator all the way to the floor - with the engine running but still cold - before they measured the peak emission levels rather than quasi-peak). Whenever I ask EMC engineers if they do this when they buy a car, they think I am crazy, and yet they see nothing wrong with doing exactly the same thing themselves when it comes to defining EMC test configurations and test methods. Claiming that a configuration more than what the regulations ask for is failing, is as ridiculous as claiming that a configuration less than what the regulations ask for is passing. Not a day goes by that I don't hear one or the other being carelessly brandished about. If you want to bring reality into the picture (usually a bad idea when it comes to law and regulations so forgive me for going there anyway), we should make sure all products which are tested at a 10 meter distance are spaced at least 10 meters apart when installed at our customer's sites. Laptops on airplanes should be spaced at least 20 rows apart (this assumes the flight attendants verify each laptop comes from a reputable manufacturer they can personally vouch for as having an EMC department populated by people of integrity). It is worthwhile to note that the ITE EMC emission
RE: Rack populating??-Rationalize it!!
I agree that the extrapolations are not, shall we say, extremely precise.
Nor are most of the measurements we do - regardless of our confidence in our
equipment and our expertise. All the more reason to try to make test
environments repeatable, and EUT's configured for maximized (worst case?)
emissions.
-Original Message-
From: Ken Javor [mailto:ken.ja...@emccompliance.com]
Sent: Friday, July 28, 2000 3:40 PM
To: CARTER; carter; 'jim_bac...@mail.monarch.com'; franz gisin;
emc-p...@ieee.org
Subject: Re: Rack populating??-Rationalize it!!
--
From: CARTER car...@amcomm.com
To: 'Ken Javor' ken.ja...@emccompliance.com, CARTER
car...@amcomm.com,
'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz Gisin
emc_...@yahoo.com, emc-p...@ieee.org
Subject: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 10:14 AM
I have long forgotten the equation for computing free space transmission
loss, but it seems to me that when radiated power, frequency, and distance
are known, one can certainly predict field intensity at any other
distance.
If that is not true, we have made a large number of terrestrial microwave
transmission paths work purely by accident.
MICROWAVE LINKS WORK IN THE FAR FIELD.
Also, CISPR 22 11.2.1:
NOTE - If the field-strength measurement at 10 m cannot be made because
of
high ambient noise levels or for other reasons, measurements of Class B
EUT's may be made at a closer distance, for example 3 m. An inverse
proportionality factor of 20 dB per decade should be used to normalize the
measured data to the specified distance for determining compliance. Card
should be taken in the measurement of large EUT's at 3 m at frequencies
near
30 MHz due to near field effects.
10 METER TO 3 METER SCALING WORKS ONLY WHEN THE EUT IS SMALL WITH RESPECT TO
3 METERS, AND WHEN THE ANTENNA DIMESNIOS ARE SMALL WITH RESPECT TO 3 M. IF
YOU TRIED TO DO A 3 METER SITE ATTENUATION WITH TWO 30 MHz TUNED DIPOLES,
YOU WOULD FIND YOU NEEDED A CORRECTION FACTOR TO ACCOUNT FOR THE NEAR FIELD
EFECTS.
Someone at IEC thinks its possible. And many labs do, in fact, test at 3
meters.
THERE HAVE BEEN A LOT OF PAPERS WRITTEN SAYING THE EXTRAPOLATION IS FAR FROM
ACCURATE.
I was not suggesting that it would be wise or meaningful to do these
things.
My point is simply that it is wise to maximize the emissions from an EUT
to
insure that we are not missing emissions which may be above the limits,
and
that the end user of the equipment is not likely to inadvertently create
such a situation.
UNDERSTAND.
_\\|//_
(' O-O ')
ooO-(_)-Ooo
Mark Carter
AM Communications, Inc.
car...@amcomm.com mailto:car...@amcomm.com
Voice: 215-538-8710
Fax: 215-538-8779
-Original Message-
From: Ken Javor [mailto:ken.ja...@emccompliance.com]
Sent: Friday, July 28, 2000 2:51 PM
To: CARTER; 'jim_bac...@mail.monarch.com'; Franz Gisin;
emc-p...@ieee.org
Subject: Re: Rack populating??-Rationalize it!!
There is a very big, very important error in Mr. Carter's point number 2.
You absolutely CANNOT extrapolate from 10 m to any other distance, unless
that other distance also happens to be both in the far field of the EUT
AND
the far field of the measurement antenna. That is the reason for a 10 m
measurement in the first place. Everyone would be doing three meter
measurements if it weren't for issues related to three meters not being
far
enough away.
Extrapolating a 10 m measurement to one airline seat away is totally
impossible. And completely unnecessary: the victim protected by the 10 m
measurement is not some arbitrary gizmoid, it is a radio receiver. You
aren't supposed to be operating a radio receiver on the airplane. More to
the point, the problem with laptops and other personal electronics on a
commercial transport is not EMI to other personal electronics, but
interference with aircraft antenna-connected receivers, whose antennas are
mounted external to the aircraft.
--
From: CARTER car...@amcomm.com
To: 'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz
Gisin
emc_...@yahoo.com, emc-p...@ieee.org
Subject: RE: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 7:35 AM
Jim,
I differ in opinion on at least two counts:
1. CISPR 22 9.1 states An attempt shall be made to maximize the
disturbance
consistent with typical applications . . . etc. This can and should be
construed as creating a worst-case scenario. Worst case is always assumed
to
be within the range of operating conditions reasonably expected. If a
laptop
will have higher emissions when it is in flames, no one (at least no one
I
know) is suggesting that it be tested that way. Setting a laptop ablaze
is
not within the range of reason.
2. The intent of making measurements at 10 meters, or 3 meters, or with
an
absorbing clamp, or any of the other requirements is not to insure that
no
interference will occur at 10 meters, 3 meters
RE: Rack populating??-Rationalize it!!
I have long forgotten the equation for computing free space transmission
loss, but it seems to me that when radiated power, frequency, and distance
are known, one can certainly predict field intensity at any other distance.
If that is not true, we have made a large number of terrestrial microwave
transmission paths work purely by accident.
Also, CISPR 22 11.2.1:
NOTE - If the field-strength measurement at 10 m cannot be made because of
high ambient noise levels or for other reasons, measurements of Class B
EUT's may be made at a closer distance, for example 3 m. An inverse
proportionality factor of 20 dB per decade should be used to normalize the
measured data to the specified distance for determining compliance. Card
should be taken in the measurement of large EUT's at 3 m at frequencies near
30 MHz due to near field effects.
Someone at IEC thinks its possible. And many labs do, in fact, test at 3
meters.
I was not suggesting that it would be wise or meaningful to do these things.
My point is simply that it is wise to maximize the emissions from an EUT to
insure that we are not missing emissions which may be above the limits, and
that the end user of the equipment is not likely to inadvertently create
such a situation.
_\\|//_
(' O-O ')
ooO-(_)-Ooo
Mark Carter
AM Communications, Inc.
car...@amcomm.com mailto:car...@amcomm.com
Voice: 215-538-8710
Fax: 215-538-8779
-Original Message-
From: Ken Javor [mailto:ken.ja...@emccompliance.com]
Sent: Friday, July 28, 2000 2:51 PM
To: CARTER; 'jim_bac...@mail.monarch.com'; Franz Gisin;
emc-p...@ieee.org
Subject: Re: Rack populating??-Rationalize it!!
There is a very big, very important error in Mr. Carter's point number 2.
You absolutely CANNOT extrapolate from 10 m to any other distance, unless
that other distance also happens to be both in the far field of the EUT AND
the far field of the measurement antenna. That is the reason for a 10 m
measurement in the first place. Everyone would be doing three meter
measurements if it weren't for issues related to three meters not being far
enough away.
Extrapolating a 10 m measurement to one airline seat away is totally
impossible. And completely unnecessary: the victim protected by the 10 m
measurement is not some arbitrary gizmoid, it is a radio receiver. You
aren't supposed to be operating a radio receiver on the airplane. More to
the point, the problem with laptops and other personal electronics on a
commercial transport is not EMI to other personal electronics, but
interference with aircraft antenna-connected receivers, whose antennas are
mounted external to the aircraft.
--
From: CARTER car...@amcomm.com
To: 'jim_bac...@mail.monarch.com' jim_bac...@mail.monarch.com, Franz
Gisin
emc_...@yahoo.com, emc-p...@ieee.org
Subject: RE: RE: Rack populating??-Rationalize it!!
Date: Fri, Jul 28, 2000, 7:35 AM
Jim,
I differ in opinion on at least two counts:
1. CISPR 22 9.1 states An attempt shall be made to maximize the
disturbance
consistent with typical applications . . . etc. This can and should be
construed as creating a worst-case scenario. Worst case is always assumed
to
be within the range of operating conditions reasonably expected. If a
laptop
will have higher emissions when it is in flames, no one (at least no one I
know) is suggesting that it be tested that way. Setting a laptop ablaze is
not within the range of reason.
2. The intent of making measurements at 10 meters, or 3 meters, or with an
absorbing clamp, or any of the other requirements is not to insure that no
interference will occur at 10 meters, 3 meters, or in a clamp, but that
test
results are repeatable. If the limit is 40 dBuV/m at 10 meters, and I test
at 20 meters, will the EUT pass? Of course not. If I have test results
from
a 10 meter set-up, I can compute the effect of an emission at any other
distance from the EUT to the next airline seat, or whatever. If I wanted
to
know that. The point is that we're all on the same page that way.
If you do the math, you can compute the interfering field strength of an
emission from a device in the next airline seat, and if you know the
immunity of the gizmoid in that seat, you can pretty much be certain that
will not be any disruptive interference.
My 2 cents
-Original Message-
From: jim_bac...@mail.monarch.com [mailto:jim_bac...@mail.monarch.com]
Sent: Friday, July 28, 2000 10:36 AM
To: Franz Gisin; emc-p...@ieee.org
Subject: Re:RE: Rack populating??-Rationalize it!!
forwarding for Franz
Reply Separator
Subject:RE: Rack populating??-Rationalize it!!
Author: Franz Gisin emc_...@yahoo.com
Date: 7/27/00 6:48 PM
Worst case is not relevant when it comes to defining
EMC test configurations for ITE equipment. I do not
know of any ITE EMC test standard that specifically
states worst case must be used. Eveywhere I look I
see words like typical or representative
RE: RE: Rack populating??-Rationalize it!!
Jim, I differ in opinion on at least two counts: 1. CISPR 22 9.1 states An attempt shall be made to maximize the disturbance consistent with typical applications . . . etc. This can and should be construed as creating a worst-case scenario. Worst case is always assumed to be within the range of operating conditions reasonably expected. If a laptop will have higher emissions when it is in flames, no one (at least no one I know) is suggesting that it be tested that way. Setting a laptop ablaze is not within the range of reason. 2. The intent of making measurements at 10 meters, or 3 meters, or with an absorbing clamp, or any of the other requirements is not to insure that no interference will occur at 10 meters, 3 meters, or in a clamp, but that test results are repeatable. If the limit is 40 dBuV/m at 10 meters, and I test at 20 meters, will the EUT pass? Of course not. If I have test results from a 10 meter set-up, I can compute the effect of an emission at any other distance from the EUT to the next airline seat, or whatever. If I wanted to know that. The point is that we're all on the same page that way. If you do the math, you can compute the interfering field strength of an emission from a device in the next airline seat, and if you know the immunity of the gizmoid in that seat, you can pretty much be certain that will not be any disruptive interference. My 2 cents -Original Message- From: jim_bac...@mail.monarch.com [mailto:jim_bac...@mail.monarch.com] Sent: Friday, July 28, 2000 10:36 AM To: Franz Gisin; emc-p...@ieee.org Subject: Re:RE: Rack populating??-Rationalize it!! forwarding for Franz Reply Separator Subject:RE: Rack populating??-Rationalize it!! Author: Franz Gisin emc_...@yahoo.com List-Post: emc-pstc@listserv.ieee.org Date: 7/27/00 6:48 PM Worst case is not relevant when it comes to defining EMC test configurations for ITE equipment. I do not know of any ITE EMC test standard that specifically states worst case must be used. Eveywhere I look I see words like typical or representative or minimum. I am willing to bet that 99.9% of you, when you bought your last car, did not insist the car be at least 6 dB under the smog limits under worst case conditions before you bought it (e.g. ask the car manufacturer to load the car down with bricks, take it to a very steep hill, facing the front of the car up the hill instead of down, and then pushed the accelerator all the way to the floor - with the engine running but still cold - before they measured the peak emission levels rather than quasi-peak). Whenever I ask EMC engineers if they do this when they buy a car, they think I am crazy, and yet they see nothing wrong with doing exactly the same thing themselves when it comes to defining EMC test configurations and test methods. Claiming that a configuration more than what the regulations ask for is failing, is as ridiculous as claiming that a configuration less than what the regulations ask for is passing. Not a day goes by that I don't hear one or the other being carelessly brandished about. If you want to bring reality into the picture (usually a bad idea when it comes to law and regulations so forgive me for going there anyway), we should make sure all products which are tested at a 10 meter distance are spaced at least 10 meters apart when installed at our customer's sites. Laptops on airplanes should be spaced at least 20 rows apart (this assumes the flight attendants verify each laptop comes from a reputable manufacturer they can personally vouch for as having an EMC department populated by people of integrity). It is worthwhile to note that the ITE EMC emission standards in the US and most of the rest of the world are designed to reduce the probability of interference, not completely eliminate it - just as smog standards are not designed to prevent death should some frustrated EMC engineer decide to take his life by running his smog-compliant car inside a closed garage. How many EMC engineers do you know who are uncomfortable with the probability aspects of the regulations and decide to take it upon themselves to get as close as possible to eliminating all potential for interference by distorting the test configuration sections of the standards until they fit under their personal definition of quality? There are a lot of places where quality (mean time between failures, for example) means more to a customer (including car owners) than whether the unit was overdesigned to meet EMC and/or smog standards. I don't feel I am any less of an EMC engineer or manager for making that statement. I would like to close by saying all the comments above are meant to act as a catalyst, food for thought so to speak, rather than be an accurate rendition of how I (or any of the companies I have worked for, am working for, or will work for) perceive what this crazy EMC discipline is all about. Feel free to discuss it as you
Re: Rack populating??-Rationalize it !!
Hi again Tania, I think from a CE point of view it's clear. If a customer is told he can do it ( or conversely you don't say it can't be done ), then you have to pass, period. An Englishman's 2 cents worth;-) Derek Walton --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
RE: Rack populating??-Rationalize it !!
Whatever the standards say, the reality is that in certain circuits, the addition of more of the same type cards or cables will produce higher emissions.From personal experience, I have changed tests reports to state exactly the number of cables attached (or cards inserted) that give passing results, but fail when one more is added! Here is where your integrity comes to a test. Do you provide equipment to your customers with 16 slots available, when you can only pass with 10 slots installed?Needless to say, you will loose your customers sooner than later. Tania Grant, tgr...@lucent.com mailto:tgr...@lucent.com Lucent Technologies, Intelligent Network Unit Messaging Solutions Group -- From: chasgra...@aol.com [SMTP:chasgra...@aol.com] Sent: Tuesday, July 25, 2000 1:33 PM To: jestuc...@micron.com; wmf...@aol.com; emc-p...@ieee.org Cc: e...@emcinteg.com Subject: Re: Rack populating??-Rationalize it !! Methinks I must protest against the keep adding until no more than xdB. As near as I can tell that was put in place by the FCC and others to minimize cable bundles - I cannot see how that applies to rack systems. The fundamental truth is that - no matter how one rationalizes it - if the full system is not tested then you flat have no idea what the real emissions profile is. Come on stop rationalizing - admit that there will be systems in the field that fail - and move on. This is a byproduct of volume vs test. The higher the volume the more rationalization takes place (witness the Class B procedures) until the emissions standards will be rendered irrelevant. --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
RE: Rack populating??
Shouldn't that be do NOT increase by MORE than 2 dB? Or DO increase by LESS than 2 dB? -Original Message- From: jestuckey [mailto:jestuc...@micron.com] Sent: Tuesday, July 25, 2000 3:47 PM To: 'wmf...@aol.com'; emc-p...@ieee.org Subject: RE: Rack populating?? The statement is: Additional cards of a representative mix will be added until the emissions do not increase by less than 2 dB. JOHN E. STUCKEY EMC Engineer Micron Technology, Inc. Integrated Products Group Micron Architectures Lab 8455 West Emerald St. Boise, Idaho 83704 PH: (208) 363-5313 FX: (208) 363-5596 jestuc...@micron.com -Original Message- From: wmf...@aol.com [mailto:wmf...@aol.com] Sent: Tuesday, July 25, 2000 05:22 To: emc-p...@ieee.org Subject: Rack populating?? I think I read in this forum that - in the context of emissions testing - it isn't necessary to fully populate a rack with, say, 4 units. Was the premise that emissions don't sum arithmetically, and the addition of the second, third units doesn't significantly effect the profile? (Obviously, if the rack is hermetic to RF, the addition of any number of units does not effect the profile.) But with allowances for some leakage... Does anyone a) remeber this thread, or b) have any opinions, or c) know of a reference WmFlanigan Ameritherm Inc Scottsville, NY --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
RE: Rack populating??-Rationalize it !!
You need to concern yourself with how may shelves to install in a rack only if you are responsible for the rack and its contents. If your responsibility is limited to the shelf, don't worry about the rack. Test rack with your shelf only - at top half. For rack concerns, I only populate racks at 50% load to test, if the shelves are of identical type. My experience has shown that if you are going to have any problems with emissions or emissions margins, you will see this at half rack with identical shelves. Anything beyond this would not be cost effective. However, if the shelves are of different type, you MUST test at full rack. Regards, Bandele Jetstream Communications, Inc. badep...@jetstream.com -Original Message- From: chasgra...@aol.com [mailto:chasgra...@aol.com] Sent: Tuesday, July 25, 2000 1:33 PM To: jestuc...@micron.com; wmf...@aol.com; emc-p...@ieee.org Cc: e...@emcinteg.com Subject: Re: Rack populating??-Rationalize it !! Methinks I must protest against the keep adding until no more than xdB. As near as I can tell that was put in place by the FCC and others to minimize cable bundles - I cannot see how that applies to rack systems. The fundamental truth is that - no matter how one rationalizes it - if the full system is not tested then you flat have no idea what the real emissions profile is. Come on stop rationalizing - admit that there will be systems in the field that fail - and move on. This is a byproduct of volume vs test. The higher the volume the more rationalization takes place (witness the Class B procedures) until the emissions standards will be rendered irrelevant. --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: Rack populating??-Rationalize it !!
Methinks I must protest against the keep adding until no more than xdB. As near as I can tell that was put in place by the FCC and others to minimize cable bundles - I cannot see how that applies to rack systems. The fundamental truth is that - no matter how one rationalizes it - if the full system is not tested then you flat have no idea what the real emissions profile is. Come on stop rationalizing - admit that there will be systems in the field that fail - and move on. This is a byproduct of volume vs test. The higher the volume the more rationalization takes place (witness the Class B procedures) until the emissions standards will be rendered irrelevant. --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
RE: Rack populating??
The statement is: Additional cards of a representative mix will be added until the emissions do not increase by less than 2 dB. JOHN E. STUCKEY EMC Engineer Micron Technology, Inc. Integrated Products Group Micron Architectures Lab 8455 West Emerald St. Boise, Idaho 83704 PH: (208) 363-5313 FX: (208) 363-5596 jestuc...@micron.com -Original Message- From: wmf...@aol.com [mailto:wmf...@aol.com] Sent: Tuesday, July 25, 2000 05:22 To: emc-p...@ieee.org Subject: Rack populating?? I think I read in this forum that - in the context of emissions testing - it isn't necessary to fully populate a rack with, say, 4 units. Was the premise that emissions don't sum arithmetically, and the addition of the second, third units doesn't significantly effect the profile? (Obviously, if the rack is hermetic to RF, the addition of any number of units does not effect the profile.) But with allowances for some leakage... Does anyone a) remeber this thread, or b) have any opinions, or c) know of a reference WmFlanigan Ameritherm Inc Scottsville, NY --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
Re: Rack populating??
In an experiment we performed using a 16 slot rack we determined that the noise floor level increased with the number of cards we added to the system. But we also observed that the spectral line associated with the primary clocks on eack card did not increas significantly as more cards were added. We finally ran the test with 4 cards as a good compromise. Derek. --- This message is from the IEEE EMC Society Product Safety Technical Committee emc-pstc discussion list. To cancel your subscription, send mail to: majord...@ieee.org with the single line: unsubscribe emc-pstc For help, send mail to the list administrators: Jim Bacher: jim_bac...@mail.monarch.com Michael Garretson:pstc_ad...@garretson.org For policy questions, send mail to: Richard Nute: ri...@ieee.org
